Adaptive multi-modal active noise control

Abstract

Low frequency Active Noise Control (ANC) has been found to work well for sound suppression in acoustic ducts. Feedforward and feedback are the two distinct methods in which to implement (ANC) schemes. A feedforward implementation is realized by utilizing a known disturbance measurement (e.g. with a microphone), and using the measurement to compute the actuator (loudspeaker) signal. Feedback control modifies dynamics of an enclosed sound field to add damping. For example, one can use a resonant filter in the feedback path that is tuned to a selected mode in the enclosed sound field to be controlled. The resonant filter model is similar to those used for positive position feedback (PPF) in structures. A method of compensating the strong speaker dynamics (phase angle) was recently investigated by Bisnette and Vipperman (2004). The method was found to improve the performance and stability of the controller. Here, the above mentioned procedure is expounded upon to provide improved multi-modal performance, using higher order band-bass filters. Methods for adapting the controller parameters (e.g., gain, frequency, bandwidth) are also presented.